/fastrl

FASTRL is a a set of benchmark tasks used in surgical training which are adapted to the reinforcement learning setting for the purpose of training agents capable of providing assistance to the surgical trainees. The benchmark is provided with the purpose of exploring the domain of human-centric teaching agents within the learning-to-teach formalism

Primary LanguagePython

Fundamentals of Arthroscopic Surgery Training: a reinforcement learning exploration and benchmark

Ivan Ovinnikov, Ami Beuret, Flavia Cavaliere, Joachim M. Buhmann

Paper | Project Website

This repository contains RL environments and corresponding config files used to train agents as stated in the paper.

If you find this repository is useful in your research, please cite the paper:

@Article{Beuret2024,
author={Beuret, Ami
and Ovinnikov, Ivan
and Cavaliere, Flavia
and Buhmann, Joachim M.},
title={Fundamentals of Arthroscopic Surgery Training and beyond: a reinforcement learning exploration and benchmark},
journal={International Journal of Computer Assisted Radiology and Surgery},
year={2024},
month={Apr},
day={29},
abstract={This work presents FASTRL, a benchmark set of instrument manipulation tasks adapted to the domain of reinforcement learning and used in simulated surgical training. This benchmark enables and supports the design and training of human-centric reinforcement learning agents which assist and evaluate human trainees in surgical practice.},
issn={1861-6429},
doi={10.1007/s11548-024-03116-z},
url={https://doi.org/10.1007/s11548-024-03116-z}
}